In order to achieve true color vision, animals need to compare the outputs of photoreceptors that are sensitive to different colors. Humans are trichromats who use Blue (S), Green (M) and Red (L) cone photoreceptors in the fovea to discriminate colors. In Drosophila, comparison occurs between the two inner photoreceptors, R7 and R8, which point in the same direction in each ommatidium (unit eye). In fact, the fly retina is composed of two classes of distinct ommatidia, which are distributed randomly throughout the retina and specialize in discriminating among different colors. 30% of ommatidia are p, which contain a UV-sensitive Rhodopsin photopigment (Rh3) in the R7 cell and blue-Rh5 in R8 and can best discriminate among short wavelengths. In the remaining 70% y ommatidia, R7 contains UV-Rh4 while R8 contains green-Rh6;they specialize in the discrimination of longer wavelengths. Because Rh3 is always associated with Rh5 and Rh4 with Rh6, this indicates that there is a signal between R7 and R8 that coordinates Rhodopsin expression. This proposal offers to study a three-step pathway that leads to the precise specification of the two classes of p and y R8. First, a BMP/Dpp signal from pR7 instructs pR8 to express Rh5. The signal is mediated by the receptor for another related growth control pathway, the Activin receptor, suggesting that the BMP and Activin pathways cross-interact. Second, a bi-stable loop insures that an unambiguous decision is made to express either Rh6 (y) or Rh5 (p). Interestingly, this loop involves the Hippo/Warts tumor suppressor pathway and the growth regulator Melted, which is part of the Insulin/TOR pathway. Thus, these tumor suppressor pathways are re-utilized in post-mitotic photoreceptors after they are no longer needed to control proliferation. Finally, maintenance of the two distinct fates is mediated by the Rhodopsin molecules themselves that act to avoid co-expression of Rhodopsins of different sensitivity. This proposal offers to study the non-canonical interaction between the BMP/Dpp and Activin pathways. It will also dissect the tumor suppressor pathway that stabilizes the decision in R8 and the involvement of Rhodopsins in maintaining the exclusive expression of one Rhodopsin gene per photoreceptor. Finally, a genetic screen using RNAi lines will identify the transcription factors that mediate the function of these signaling pathways and allow precise expression of Rhodopsins. This work has clear relevance to our understanding of retinal patterning but also of the Hippo/Warts tumor suppressor pathway whose upstream components and transcriptional effectors remain unknown. Genetics offers the best chance to identify these new components. PUBLIC HEALTH RELEVANCE: This project addresses several fundamental questions, not only about retinal patterning but also about the function of tumor suppressor and growth control pathways. We anticipate that we will be able to provide critical identification of upstream and downstream components of these pathways, in particular their output at the level of gene expression, as well as an understanding of their cross-interactions.